Openi CEO Sam Altman surprised the tech world Thursday with the release of Operator, his company’s first AI agent that can act autonomously on the Web.
But Openai isn’t Altman’s only venture that’s trying to capitalize on the popularity of AI agents.
Altman’s global project now wants to create tools that connect certain AI agents to people’s online personas, allowing other users to verify that an agent is acting on behalf of a person, according to its Chief Product Officer, Tiago Sada.
World, a Web3 project by Altman and Alex Blania’s Tools for Humanity that was previously known as Worldcoin, is based on the idea that it will eventually be impossible to distinguish humans from AI agents on the Internet. To address this, World wants to create “Prove of Human” digital tools. After scanning your eyeball with a silver metal ball, the world will give you a unique identifier on the blockchain to verify that you are a human.
It’s worth noting that Tools for Humanity have a reputation for pivoting and getting into controversies. The world began as a crypto project, was temporarily banned in some countries, and was recently ordered to allow Europeans to delete their biometric data upon request. In October, the project dropped “coins” from its name, signaling that it would focus more on human verification and less on cryptocurrency.
While verifying AI agents so they can do things on behalf of people may seem like a divergence from the worldwide mission to verify humans with Web3 IDs, Sada says it’s a critical and logical next step for the project.
“This idea of delegating your ‘proof of personality’ to an agent and letting them act on your behalf is actually very important,” he said in an interview with TechCrunch. “Instead of only allowing people you think are human (on your website), you will also allow AI agents that represent a real human being. This is where World ID comes in. “
World ID technology could also be used to license AI agents to act on your behalf, Sada said. In a recent blog post, the project notes that its proof of human tools will not only distinguish humans from robots, as they do today, but could help people control a network of online AI agents.
The world appears to be expanding its tools to verify “digital actions” that belong to a person, even if they didn’t come from their account. The project says its AI tools will prove essential in 2025, but Sada said the company “will have to see” whether any of those tools involve connecting AI agents to people. It certainly seems like the startup is thinking about it.
“There are some apps where it doesn’t matter if a real person is using it or an agent acting on their behalf. You just care about knowing that there’s a person supporting that interaction,” Sada said.
Many websites today use tools from CloudFlare and Snowflake to block AI bots from scraping their website. Early adopters of the Openi operator have found that some websites have blocked Openi’s new agent by default.
But Sada says companies may reconsider their permissions to allow some AI agents to use their services.
“Ultimately, companies want to sell more. They want to serve more users,” Sada said. “If a delivery app’s sales increase because there’s a lot of agents coming in and responding on behalf of their users, the company will be happy.”
That example is quite timely. Openai just announced partnerships with Uber, Instacart and Doordash to allow its new operator agent to use their platforms. This is a big change for these companies, which are used to serving users on their apps. Instead, online businesses now appear to be preparing for a future in which some of their user interactions will be facilitated by AI agents.
“If an agent comes in with a blue check mark and says, ‘Yes, I’m a bot. Don’t worry, don’t go crazy. I’m actually here to buy a hot dog for Tiago, “then those companies can let him (agent) come in,” Sada said.
The problem with the entry of every AI agent, or every bot on the Internet, is that it could open your website to attacks, scams, or other bad actors. By allowing in just a couple of AI agents per person, Sada says, companies can access new users through these agents without risking the integrity of their site.
Altman’s many initiatives may seem random, but seen together, they create a network of tools to power a future that has artificial intelligence, and by extension, Openai, at its center. His nuclear fusion startup, Energy Helion, could one day provide power to Openai’s data centers, and the longevity science startup he backs, retro biosciences, is already using Openi’s models to try to increase human lifespan .
A key question around the world is whether its tools will one day integrate more deeply with Openi. Altman speaks at world events, supposedly talks to the team every week, and the project continues to lean on his AI work.
Platforms appear to be lining up to let Openi agents use their services, but the world could play a significant role in vetting those agents and others.
TechCrunch has a newsletter focused on AI! Sign up here to get it delivered to your inbox every Wednesday.
